there is no dedicated node that extracts how close two terms are mentioned in a sentence. One workaround to do this (if you don't want to implement your own node) is to extract the sentences containing both terms as strings. Cut off the substring before the first term and the substring after the last term and count the number of whitespaces in between.